1️⃣ Question 1

How do generative AI models help discover new drugs?

• ❌ Analyze lifestyle factors

• ❌ Analyze medical images

• ✅ Analyze molecular structures of known medications and their impact on biological targets

• ❌ Analyze genetic information

Explanation:
Drug discovery uses generative models to design new molecules based on known molecular structures.

---

2️⃣ Question 2

Generative AI models best suited for creating synthetic images & modifying attributes:

• ❌ Medical image analysis

• ✅ Creating unique designs & enhancing creative workflows in fashion

• ❌ Predicting retail demand

• ❌ Manufacturing cost optimization

Explanation:
GANs & diffusion models excel at visual generation & style manipulation, used heavily in fashion.

---

3️⃣ Question 3

Generative models with two neural networks:

• ❌ Flow-based models

• ❌ VAEs

• ❌ Autoregressive

• ✅ GANs (Generative Adversarial Networks)

Explanation:
GANs contain a generator and discriminator trained against each other.

---

4️⃣ Question 4

How do generative AI models help manage financial risks?

• ✅ Simulating scenarios such as market crashes or economic downturns

• ❌ Analyze customer behavior

• ❌ Analyze financial data

• ❌ Detect anomalies

Explanation:
Generative models can simulate rare but high-impact financial events.

---

5️⃣ Question 5

Which generative models compress data effectively?

• ❌ GANs

• ✅ VAEs (Variational Autoencoders)

• ❌ Flow-based models

• ❌ Autoregressive models

Explanation:
VAEs learn a latent representation that compresses data while retaining information.

---

6️⃣ Question 6

How does generative AI help augment data?

• ❌ Generate software code

• ✅ Create synthetic data that mimics real data

• ❌ Uncover hidden patterns

• ❌ Generate business reports

---

7️⃣ Question 7

How do generative AI models detect outliers?

• ❌ Use latent code

• ❌ Graph representation

• ❌ Natural language understanding

• ✅ Learn boundaries of the normal data distribution

Explanation:
Outliers lie outside the learned distribution.

---

8️⃣ Question 8

Which models are “sequential data champions”?

• ❌ Flow-based

• ✅ Autoregressive models

• ❌ GANs

• ❌ VAEs

Explanation:
Autoregressive models predict one step at a time → perfect for sequences.

---

9️⃣ Question 9

How do generative AI models handle missing values?

• ❌ Graph representations

• ❌ Latent code only

• ✅ Learn underlying patterns & generate plausible missing values

• ❌ Natural language SQL

---

🔟 Question 10

Why are flow-based models efficient at sampling?

• ❌ Predict future trends

• ❌ Generate multimodal data

• ✅ Perform direct modeling of the probability distribution of the data

• ❌ Compress data

Explanation:
Flow-based models use invertible transformations that allow exact likelihood computation & fast sampling.

---

🧾 Summary Table